Inhibitors of the complement system currently in development for cardiovascular disease.

Controlled activation of the complement system is critical to the host-defense response of the immune system. Activated complement is responsible for the stimulation of a localized protective inflammatory response to either invading microorganisms or foreign molecules (toxins). However, the autologous activation of the complement system can have devastating consequences on many organ systems. This review discusses the various pathways involved in the activation of the complement system and the multiple levels of control established within the body to regulate activation. It also focuses on the role of complement activation in cardiovascular disease, especially myocardial ischemia and reperfusion injury as well as in cardiopulmonary bypass procedures. Lastly, this review also provides a comprehensive overview of both biologically derived proteins and chemically developed inhibitors of the complement system that range from those that are currently in the discovery stage to those that are in clinical development as novel therapeutic agents.

Recombinant bactericidal/permeability-increasing protein (rBPI21) for treatment of parvovirus enteritis: a randomized, double-blinded, placebo-controlled trial.

We evaluated the ability of an antimicrobial and endotoxin-neutralizing agent, the recombinant amino terminal fragment of bactericidal permeability-increasing protein (rBPI21), to decrease plasma endotoxin concentration and severity of clinical signs of canine parvovirus and to improve survival. This randomized, double-blinded, placebo-controlled clinical trial included 40 client-owned dogs and 9 normal puppies from a closed research colony. Dogs weighing >5 kg (11 lb) with fecal antigen-confirmed parvovirus and clinical signs of vomiting and diarrhea were randomly assigned to receive placebo or rBPI21 infusion over 6 hours. Plasma endotoxin concentration was measured at 0, 3, and 6 hours of infusion. Owners chose continued medical care with either the Veterinary Hospital of the University of Pennsylvania Internal Medicine Service or a local veterinarian. Telephone follow-up was conducted at 14 days. Surviving dogs were reevaluated at >30 days (recovered group), at which time plasma samples for measurement of endotoxin concentration were obtained. Plasma endotoxin concentrations were significantly higher in dogs with parvovirus than in normal or recovered dogs. Despite 90% survival, the rBPI21 treatment did not have a significant effect on outcome, duration of hospitalization, or plasma endotoxin concentrations. Treatment in a tertiary care hospital, however, significantly improved survival but resulted in a significantly increased duration of hospitalization. Endotoxemia occurs in dogs with parvovirus enteritis, but rBPI21 is not associated with improved survival.

Synergistic effect of a recombinant N-terminal fragment of bactericidal/permeability-increasing protein and cefamandole in treatment of rabbit gram-negative sepsis.

As a consequence of their bactericidal actions, many antibiotics cause the release of endotoxin, a primary mediator of gram-negative sepsis. Bactericidal/permeability-increasing protein (BPI) has bactericidal activity and neutralizes endotoxin in vitro and in vivo. We sought to examine the effect of a recombinant N-terminal fragment of BPI (rBPI21) in conjunction with cefamandole, a cephalosporin antibiotic, in the treatment of Escherichia coli bacteremia and septic shock in rabbits. Cefamandole (100 mg/kg of body weight) was injected intravenously. This was followed by simultaneous 10-min infusions of E. coli O7:K1 (9 x 10(9) CFU/kg) and rBPI21 (10 mg/kg). rBPI21 was continuously infused for an additional 110 min at 10 mg/kg/h. The administration of rBPI21 in conjunction with the administration of cefamandole prevented the cefamandole-induced increase of free endotoxin in plasma, accelerated bacterial clearance, ameliorated cardiopulmonary dysfunction, and thereby, prevented death, whereas neither agent alone was protective in this animal model. The efficacy of the combined treatment with rBPI21 and cefamandole suggests a synergistic interaction between the two agents. The data indicate that rBPI21 may be useful in conjunction with traditional antibiotic therapy.

rBPI23 attenuates endotoxin-induced cardiovascular depression in awake rabbits.

We determined the effect of a recombinant N-terminal fragment of bactericidal/permeability-increasing protein (rBPI23) on hemodynamic and renal sympathetic responses to lethal endotoxemia in unanesthetized rabbits. Endotoxin was continuously infused intravenously (200 micrograms/kg/h) over 120 min with simultaneous infusion of either rBPI23 (3 mg/kg bolus followed by 6 mg/kg/h over 120 min; n = 6) or thaumatin (the same dose as rBPI23), a control cationic protein with a molecular weight and isoelectric point similar to that of rBPI23 (n = 9). Tissue blood flow was also determined using colored microspheres to the left ventricle, renal cortex, liver, and skeletal muscle. Seven of nine animals treated with endotoxin and thaumatin died between 45 and 120 min after start of the infusion, whereas all animals with rBPI23 treatment were alive throughout the entire 2 h experimental period. A transient increase in renal sympathetic nerve activity was observed in the thaumatin-treated animals followed by sympathoinhibition with concomitant decreases in heart rate, blood pressure, and cardiac output. Tissue blood flow to all measured organs gradually decreased in animals receiving endotoxin and thaumatin. However, rBPI23 abolished all these deleterious responses to endotoxin. In conclusion, rBPI23 attenuates the acute lethal sympathoinhibitory and hemodynamic effects of endotoxemia in awake rabbits.

Protective effects of an N-terminal fragment of bactericidal/permeability-increasing protein in rodent models of gram-negative sepsis: role of bactericidal properties.

Effects of an N-terminal fragment of bactericidal/permeability increasing protein (rBPI21) on bacterial infections were determined. Intravenous (i.v.) rBPI21 increased survival and reduced bacteremia in rats after an iv injection of Escherichia coli O7:K1 bacteria. rBPI21 inhibited the rise in tumor necrosis factor-alpha resulting from challenge with 2 strains of E. coli. Intraperitoneal (ip) injection of rBPI21 increased survival of mice after ip injection of E. coli O7:K1 and Pseudomonas aeruginosa and reduced bacteria in peritoneal lavage fluid and blood and inhibited cytokine production in response to E. coli. rBPI21 alone did not protect mice challenged with E. coli O111:B4 but was protective and reduced bacterial counts when administered in combination with the antibiotic cefamandole. The data show that protection with rBPI21 is associated with reductions in bacterial counts and is enhanced by antibiotics. Bactericidal activity, in addition to antiendotoxin activity, is involved in the efficacy of rBPI21 in models of gram-negative infection.

Protective effects of a recombinant N-terminal fragment of bactericidal/permeability increasing protein on endotoxic shock in conscious rabbits.

Endotoxin (lipopolysaccharide, LPS) can induce shock, multiple organ failure, and death. A recombinant N-terminal fragment of bactericidal/permeability increasing protein, rBPI23, binds with high affinity to gram-negative bacterial LPS and neutralizes its biological activity. We sought to determine the effect of rBPI23 on LPS-induced respiratory dysfunction and cardiovascular depression in conscious rabbits. Rabbits were injected with Escherichia coli O113 LPS (6 micrograms/kg) and treated with rBPI23 (2 mg/kg), vehicle, or control protein after recovery from surgery performed to implant catheters for hemodynamic assessments and intravenous injections. LPS challenge caused respiratory dysfunction including tachypnea, significant decreases in arterial O2 tension (PO2), arterial oxygen content, and an increase in alveolar-arterial O2 gradient (A-aDO2). LPS administration also resulted in profound and prolonged decreases in mean arterial blood pressure and cardiac index. Treatment with rBPI23 prevented LPS-induced respiratory dysfunction and significantly ameliorated the cardiovascular depression. 5 of 16 LPS-challenged animals died of respiratory failure and acidosis, whereas none died in the rBPI23 treated group (p = .11). The results demonstrate that rBPI23 protects animals against LPS-induced cardiopulmonary depression in endotoxic shock.

Protective effect of a recombinant fragment of bactericidal/permeability increasing protein against carbohydrate dyshomeostasis and tumor necrosis factor-alpha elevation in rat endotoxemia.

Endotoxin (lipopolysaccharide, LPS), a component of the gram-negative bacterial cell wall, induces carbohydrate dyshomeostasis and the release of proinflammatory cytokines such as tumor necrosis factor-alpha (TNF-alpha) when administered to experimental animals. Bactericidal/permeability increasing protein (BPI), a cationic protein found in human neutrophil granules, binds with high affinity to LPS and is capable of neutralizing its biological activity. The present study was designed to determine if a recombinant N-terminal fragment of BPI, rBPI23, attenuates LPS-induced alterations in serum glucose, lactate, and TNF-alpha in rats. In anesthetized animals challenged with a 30 min infusion of Escherichia coli O111:B4 LPS (0.25 mg/kg), there was an early transient increase in serum levels of glucose followed by a drop to 60% of those found in saline control rats. A prolonged elevation in serum levels of lactate and a transient, but marked, elevation of TNF-alpha were also observed following LPS infusion. These LPS-induced changes were inhibited significantly by simultaneous infusion of rBPI23. Different dose-response profiles of rBPI23 on LPS-induced alterations in glucose, lactate and TNF-alpha were observed. When rBPI23 was infused 30 min after the initiation of LPS infusion, it significantly inhibited the alterations in glucose and lactate, but not TNF-alpha. The rise in TNF-alpha was reduced significantly with a 15 min delayed infusion of rBPI23. A control protein failed to alter any responses to LPS. The results indicate that rBPI23 can provide significant protection against the metabolic disturbances and TNF-alpha release associated with endotoxemia. In addition, the results suggest that LPS-induced metabolic alterations in glucose and lactate are at least partially independent of TNF-alpha release.

The bactericidal/permeability increasing protein of neutrophils is a potent antibacterial and anti-endotoxin agent in vitro and in vivo.

The Bactericidal/Permeability Increasing protein (BPI) is a major constituent of the azurophilic granules of human and rabbit polymorphonuclear leukocytes (PMN). The cDNA of the highly conserved protein has been isolated from man, rabbit and cow. The ca. 50 kDa BPI and a ca. 25 kDa bioactive N-terminal fragment are cytotoxic only for Gram-negative bacteria (GNB). This target-cell specificity reflects the strong attraction of the highly cationic protein for the negatively charged lipopolysaccharides (LPS) in the bacterial envelope. Native and recombinant (r) holo-BPI and the N-terminal fragment (rBPI-23) bind with high affinity (apparent Kd 1-10 nM) to all forms of isolated LPS so far examined, and inhibit the numerous biological effects of LPS in vitro (including in whole blood ex vivo) as well as in animals. Under the same conditions the antibacterial activities of holo-BPI and rBPI-23 against GNB with rough chemotype LPS are the same, but against serum-resistant and smooth chemotype GNB rBPI-23 is up to 30-fold more potent than holo-BPI. Holo-BPI and rBPI-23 protect mice, rats and rabbits against lethal cytotoxic effects of LPS and in some cases against lethal inoculations with live GNB.

Tonic descending modulation of spinal neuronal responses to activation of renal receptors.

Experiments were conducted in anesthetized cats to determine if spinal neuronal responses to activation of renal receptors are tonically modulated by descending spinal pathways. Eighty-seven thoracolumbar spinal neurons with renal and somatic input were tested for responses to occlusion of the renal vein, renal artery, and ureter before, during, and after cooling the spinal cord rostral to the recording site. Cooling increased the number of neurons that responded as well as the magnitude of the responses to renal vein (RVO), renal artery (RAO), and ureteral occlusion (UO). RVO increased cell activity of 21 neurons from 12.5 +/- 2.7 to 31.7 +/- 6.0 spikes/s during cooling. UO increased cell activity of 24 neurons from 9.0 +/- 2.1 before cooling to 25.0 +/- 4.9 spikes/s during cooling. Cold block increased the magnitude of both types of responses to RAO that were due to mechanical deformation of the renal artery and prolonged renal ischemia. These data show that the majority of spinal neuronal responses to renal receptor stimulation are modulated by tonic inhibitory influences. Thus these results provide a mechanism by which the brain may control spinal circuitry that underlies reflexes of renal origin.

Protective effect of a recombinant amino-terminal fragment of bactericidal/permeability-increasing protein in experimental endotoxemia.

Bactericidal/permeability-increasing protein (BPI), a cationic protein found in neutrophil granules, binds with high affinity to gram-negative bacterial lipopolysaccharide (LPS) and can inhibit its actions in vitro. The in vivo efficacy of a recombinant 23-kDa amino-terminal LPS-binding fragment of BPI (rBPI23) was assessed in a mouse model of lethal endotoxemia. Systemic administration of rBPI23 protected actinomycin D-sensitized mice from lethal LPS (Escherichia coli O111:B4) challenge in a dose-dependent manner, with almost complete protection at the highest dose (10 mg/kg; 93% survival vs. 13% in vehicle-treated controls). Surviving rBPI23-treated animals did not show histopathologic signs of tissue damage evident in control animals that had died after LPS challenge. rBPI23 also attenuated the LPS-induced elevation in serum levels of tumor necrosis factor-alpha and interleukin-1 alpha, mediators believed to be involved in the pathogenesis of endotoxemia and sepsis. Thus, rBPI23 may be a potential new therapeutic agent for the treatment of gram-negative bacterial infection and sepsis.

Recombinant amino terminal fragment of bactericidal/permeability-increasing protein prevents hemodynamic responses to endotoxin.

We sought to determine if a recombinant amino terminal fragment of bactericidal/permeability increasing protein (rBPI23) alters the hemodynamic responses to endotoxin. Experiments were performed on Sprague Dawley rats anesthetized with Ketamine and xylazine. In rats challenged with a 30 min infusion of 0.25 mg/kg lipopolysaccharide (LPS; Escherichia coli 0111:B4), there were early (30-90 min), significant increases in cardiac index, heart rate, and stroke volume, accompanied by significant decreases in blood pressure and total peripheral resistance. For the remainder of the 210 min observation period, cardiac index, and stroke volume progressively declined to levels significantly below those of control rats receiving only vehicles. At the same time, blood pressure and total peripheral resistance steadily increased above the vehicle control group. Infusion of 3 mg/kg of rBPI23 abolished these LPS-induced hemodynamic responses. A dose of 1.0 mg/kg of rBPI23 was associated with a modest, significant inhibition of changes evoked by LPS, whereas 0.3 mg/kg was without significant effect. Thaumatin, a control cationic protein with molecular weight and isoelectric point similar to those of rBPI23, failed to alter any responses to LPS. These results indicate that rBPI23 produces a dose-dependent inhibition of hemodynamic changes, associated with endotoxemia, and provides further support for the potential utility of rBPI23 as a therapeutic agent in the treatment of gram-negative sepsis and infection.

Effects of renal receptor stimulation on neurons within the ventrolateral medulla of the cat.

Experiments were performed to determine if activation of renal receptors by occlusion of the renal artery, renal vein, or ureter would alter activity of cells within the ventrolateral medulla of the cat. Extracellular unit recordings were obtained from 195 cells located within the rostral ventrolateral medulla of 90 alpha-chloralose-anesthetized cats. Fifty-five of 195 cells (28.2%) tested for responses to renal receptor activation responded to at least one of the occlusions. Occlusion of the ureter increased the activity of 25 cells from 9.7 +/- 3.7 to 23.0 +/- 6.5 impulses/s and decreased the activity of 5 cells from 11.9 +/- 3.6 to 3.5 +/- 1.2 impulses/s. Occlusion of the renal vein increased the activity of seven cells from 7.5 +/- 3.3 to 22.3 +/- 7.3 impulses/s and decreased the activity of six cells from 13.8 +/- 3.8 to 4.1 +/- 2.0 impulses/s. Renal artery occlusion elicited solely excitatory responses from 43 cells. Thirty-one of the 43 cells increased their activity within 0-3 s of the onset of renal artery occlusion from 4.1 +/- 0.8 to 12.6 +/- 1.2 impulses/s. Renal artery occlusion increased the activity of 10 out of 43 cells with a mean latency of 26.1 +/- 6.5 s from 8.3 +/- 2.5 to 29.6 +/- 9.3 impulses/s. Twenty-four of the 55 (43.6%) responders were responsive to two or more forms of renal receptor activation. These results demonstrate that activation of renal mechanoreceptors and chemoreceptors affects cells within the ventrolateral medulla of the cat.(ABSTRACT TRUNCATED AT 250 WORDS)

Renal afferent input to the ventrolateral medulla of the cat.

Experiments were performed to determine if information from the kidneys projects to the rostral ventrolateral medulla. Extracellular action potentials were recorded from 148 cells within the rostral ventrolateral medulla of alpha-chloralose-anesthetized cats. Cells within the rostral ventrolateral medulla were tested for responses to electrical stimulation of both left and right renal nerves. Electrical stimulation of renal nerves excited 144 cells (97.3%) and inhibited 4. The majority of cells received either bilateral or contralateral renal nerve input. Cells with bilateral renal nerve input responded to contralateral renal nerve stimulation with a significantly greater number of impulses compared with ipsilateral renal nerve stimulation (P less than 0.05). All cells but one responding to renal nerve stimulation had convergent somatic input. Comparisons between thresholds for cell responses and activation thresholds for the A and C volleys of the compound action potential recorded in the least splanchnic nerve revealed that 44 cells required activation of A delta-fibers, and 12 cells required activation of both A delta- and C-fibers. A conditioning stimulus applied to renal nerves on one side significantly decreased the response elicited by a test stimulus applied to the renal nerves on the opposite side for at least 300 ms (P less than 0.05). The demonstration that an afferent connection exists between the kidneys and the ventrolateral medulla suggests that the rostral ventrolateral medulla may play a role in mediating supraspinal reflexes of renal origin.

Tonic descending modulation of spinal neurons with renal input.

Experiments were conducted to determine the influence of tonically active descending pathways on thoracolumbar spinal neurons that respond to renal nerve stimulation in anesthetized cats. We examined the effect of reversible blockade of spinal conduction on spontaneous activity, responses to renal nerve stimulation and responses to somatic stimuli of 71 spinal neurons. Mid-thoracic cold block resulted in enhanced responses (tonically inhibited neurons), reduced responses (tonically excited neurons), or did not affect neuronal responses. The spontaneous activity of 47 of 69 neurons (68%) increased from 7.3 +/- 2.0 spikes/s before cooling to 23.3 +/- 4.5 spikes/s during cooling. Activity of 8 neurons (12%) decreased while 14 (20%) had no change in activity. Cooling increased the responses of 51 of 71 neurons (72%) to renal nerve stimulation. Renal nerve stimulation evoked a two-fold increase in both short latency (early) and long latency (late) responses. Four neurons had a late response which was revealed by cold block. Cooling decreased responses of 8 of 71 neurons (11%) and 9 neurons (13%) were not affected. Cooling increased the early responses but decreased the late responses of 3 of 71 neurons (4%). All neurons had somatic receptive fields and 33 of 56 exhibited increased responses to somatic stimulation during cooling. In addition, receptive field sizes of 26 neurons increased. Four neurons had a decrease and 25 neurons had no change in receptive field size during cooling. These data indicate that tonically active descending pathways modulate the activity of most spinal neurons with renal input and the major effect of these pathways is inhibitory. This influence may be important in the modulation of spinal circuits that participate in reflexes evoked by renal afferent fibers.

Bowditch Lecture. Renal afferent inputs to ascending spinal pathways.

Studies of renal afferent fibers and their functions have continued since the work of Pines in 1959 (Fiziol. Zh. SSSR Im. I M Sechenova 45: 1339-1347, 1959). The kidney contains mechanoreceptors and chemoreceptors that appear to have two major functions. First, renal mechano- and chemoreceptors evoke a variety of renorenal reflexes, while more global cardiovascular reflexes are primarily evoked by renal mechanoreceptors. A second function of renal afferent fibers is to cause the pain of renal disease. Recent studies suggest that renal afferent fibers may also regulate secretion of vasopressin from the pituitary gland. Substantial evidence indicates that, although most renal afferent fibers enter the spinal cord, their functions depend to a large extent on supraspinal circuitry. Thus our research has focused on defining characteristics of spinal neurons that relay renal information to the brain. In the cat, neurons in the L2-T11 segments with excitatory responses to renal A delta and C fiber input project to the medial medullary reticular formation and to the caudal and rostral ventrolateral medulla. Renal afferent information reaches these cells by way of the least splanchnic nerve and by way of more than one dorsal root. In the monkey spinothalamic neurons in the L3-T10 segments respond to renal nerve stimulation. Excitatory responses predominate, but inhibitory responses occur in L2 and L3. These cells also respond to renal A delta and C fibers. Stimulation of renal mechanoreceptors by occlusion of the ureteropelvic junction or renal vein excites feline spinoreticular neurons. Graded increases in renal vein pressure produce graded increases in cell responses. Activation of renal chemoreceptors increases activity of spinal interneurons. Within the L2-T11 segments, cells responding to ureteral occlusion are located caudally, cells with responses to renal artery occlusion are located rostrally, and cells responding to renal vein occlusion are located in between. The differential locations of cells with these inputs suggests the existence of a coding mechanism for different renal receptor populations. Distention of the renal pelvis is a potent stimulator of primate spinothalamic neurons. These neurons encode renal pelvic pressures in the noxious range and appear to be important in mechanisms of renal pain.

Convergence of phrenic and cardiopulmonary spinal afferent information on cervical and thoracic spinothalamic tract neurons in the monkey: implications for referred pain from the diaphragm and heart.

1. Spinothalamic tract (STT) neurons in the C3-T6 spinal segments were studied for their responses to stimulation of phrenic and cardiopulmonary spinal afferent fibers. A total of 142 STT neurons were studied in 44 anesthetized, paralyzed monkeys (Macaca fascicularis). All neurons were antidromically activated from the ventroposterolateral nucleus and/or medial thalamus. 2. Electrical stimulation of phrenic afferent fibers (PHR) excited 43/58 (74%), inhibited 2/58 (3%), and did not affect 13/58 (13%) of cervical STT neurons. Neurons with excitatory somatic fields confined to the proximal limb or encompassing the whole limb were excited to a significantly greater extent by electrical stimulation of PHR than were neurons with somatic fields confined to the distal limb. Mechanical stimulation of PHR by probing the exposed diaphragm excited 11/22 (50%), inhibited 3/22 (14%), and did not affect 8/22 (36%) cervical STT neurons. 3. The technique of minimum afferent conduction velocity (MACV) was used to obtain information about the identity of the PHR that excited 35 cervical STT neurons. Evidence was obtained for excitation of these neurons by group II and III PHR. The mean +/- SE MACV for all neurons was 14 +/- 2 m/s. 4. Electrical stimulation of cardiopulmonary spinal afferent fibers excited 41/57 (72%), inhibited 8/57 (14%), and did not affect 8/57 (14%) of cervical STT neurons. Neurons with excitatory somatic fields confined to the proximal limb or encompassing the whole limb were excited to a significantly greater extent by electrical stimulation of cardiopulmonary spinal afferents than were neurons with somatic fields confined to the distal limb. 5. Excitatory convergence of PHR and cardiopulmonary spinal afferent input was observed for 36/57 (63%) cervical STT neurons. 6. Electrical stimulation of PHR excited 36/84 (43%), inhibited 25/84 (30%), and did not affect 23/84 (27%) of thoracic STT neurons. All of these neurons received excitatory cardiopulmonary spinal afferent input. 7. Neurons were more likely to be excited by electrical stimulation of PHR if they were located in C3-C6 spinal segments. Furthermore, the net excitatory effect of PHR input decreased in more caudal segments, such that thoracic STT neurons were weakly excited relative to cervical STT neurons. 8. We conclude that cervical STT neurons with excitatory somatic fields that include or are restricted to proximal sites are excited by electrical or mechanical stimulation of PHR. Those effects demonstrate a physiological substrate for pain referred from the diaphragm to the shoulder in patients with pleural effusions or subphrenic abscesses.(ABSTRACT TRUNCATED AT 400 WORDS)

Responses of spinoreticular cells to graded increases in renal venous pressure.

Previous work established that occlusion of the renal vein excites spinoreticular tract (SRT) neurons of the cat. The present study was designed to determine the relationship between renal vein pressure level and SRT cell activity. Experiments were performed on 40 cats that were anesthetized with alpha-chloralose. Sixty SRT neurons in the T12-L2 segments were tested for responses to a renal vein pressure (RVP) of 60 mmHg. Twenty-three cells responded with an increase in activity. Stimulus response relationship for these cells were determined over the RVP range of 10-80 mmHg. RVP thresholds averaged 25 +/- 3 mmHg. Above this level greater increases in RVP were associated with greater increases in neuronal activity. At RVP of 80 mmHg cell activity increased from 4 +/- 2 to 21 +/- 5 spikes/s. Cells with both A delta- and C-fiber renal afferent inputs had significantly greater responses and lower thresholds than cells with only A delta input. Renal vein occlusion evoked increases in blood pressure. At a renal vein pressure of 80 mmHg, pressor responses averaged 16 +/- 5 mmHg. No significant changes in heart rate were observed. The results demonstrate that SRT cells are capable of encoding the level of RVP. Such responses may be important for grading of reflexes of renal origin that require supraspinal circuitry.

Cardiopulmonary sympathetic afferent excitation of lower thoracic spinoreticular and spinothalamic neurons.

1. Responses of spinoreticular (SRT) and spinothalamic (STT) neurons located in the T7-T9 segments to cardiopulmonary sympathetic afferent (CPS) stimuli were studied in 27 cats that were anesthetized with alpha-chloralose. 2. CPS stimulation excited 32 SRT and 10 STT neurons. Each neuron was also excited by stimulation of the left greater splanchnic nerve (SPL) and had a somatic receptive field that was most commonly located on the upper abdomen and over the lower rib cage. An additional 12 SRT and 3 STT neurons received input from SPL and somatic structures but failed to respond to CPS stimulation. 3. CPS stimulation evoked early responses (23 cells) or both early and late responses (19 cells) that had average latencies of 12.7 +/- 1.8 and 88.2 +/- 13.1 (SE) ms, respectively. Latencies of responses to SPL stimulation were significantly shorter and averaged 8.1 +/- 0.9 and 46.1 +/- 7.1 ms. Magnitudes of early responses to SPL stimulation were significantly greater than responses to CPS stimulation; however, late responses were not different. 4. Responses to CPS stimulation were inhibited by a prior conditioning stimulus applied to SPL. Greatest inhibition occurred at a conditioning-test interval of 40 ms, and inhibition lasted for at least 300 ms. Inhibition of responses to SPL stimulation could be evoked by conditioning stimuli applied to CPS; however, the inhibition was significantly less than that evoked by SPL stimulation on responses to CPS stimulation. 5. Thirty-eight neurons were tested for responses to injection of bradykinin (4 micrograms/kg) into the left atrium. Discharge rate of 17 cells increased from 5 +/- 2 to 12 +/- 4 Hz. Four cells were tachyphylactic to repeated injections. Injections of bradykinin into the thoracic aorta did not significantly alter cell activity. Bilateral cervical vagotomy had no effect on responses to intracardiac bradykinin. 6. The results indicate that lower thoracic SRT and STT neurons are excited by CPS stimuli including noxious stimulation of the heart. However, comparison of these responses with previously reported responses of upper thoracic SRT and STT neurons indicate that there is a decrease in effectiveness of CPS stimuli from upper to lower thoracic segments. Convergence of CPS and abdominal inputs onto lower thoracic pain pathways could explain abdominal pain that is occasionally associated with cardiac disease.

Cardiopulmonary sympathetic afferent input to lower thoracic spinal neurons.

Spinal neuronal responses to stimulation of cardiopulmonary sympathetic afferent (CPS) fibers were studied in 25 alpha-chloralose-anesthetized cats. Eighty-two neurons located in the T7-T9 segments were tested for responses to electrical stimulation of CPS fibers. Activity of 55 neurons was altered; 37 were excited, 10 were inhibited, and 8 were both excited and inhibited. All 55 cells with CPS input also responded to stimulation of somatic receptors and the left greater splanchnic nerve (SPL). Somatic receptive fields were primarily located on the upper portion of the abdomen and left lower rib cage. Short and long latency responses occurred following CPS and SPL stimulation. Latencies of responses to CPS stimulation were significantly longer than latencies of responses to SPL stimulation (P less than 0.05). Early responses to CPS stimulation were significantly less in magnitude compared to early responses to SPL stimulation (P less than 0.05). Cell responses to CPS stimulation were reduced in magnitude for as long as 300 ms when a conditioning stimulus was applied to SPL. Inhibitory responses of 10 cells to CPS fiber stimulation were best observed during repetitive stimulation. Eight of the cells were also inhibited by repetitive stimulation of SPL. Injection of bradykinin (4 micrograms/kg) into the left atrium increased activity of 16/30 cells from 8 +/- 2 to 22 +/- 5 spikes/s. The results demonstrate that CPS fiber stimulation alters activity of lower thoracic spinal neurons but not as intensely as SPL stimulation. These neurons may participate in cardiac-abdominal visceral reflexes or the pain of cardiac origin that is referred to the abdomen.